(Georgia Institute of Technology, 2022-08-22)
Lombardo, Sarah F.
The need for novel, super-high K dielectric gate oxides has substantially increased in recent years. With the equivalent oxide thickness in advanced nodes reaching a limit, materials innovation can enable increased dielectric constants in gate oxide stacks beyond the current limit, providing significant enhancement in logic technologies. Capacitance enhancement and super-high K dielectric gate stacks require the use of FE materials to be stabilized in an otherwise unstable state, resulting in an effective static negative capacitance (NC). However, the structural ferroelectric pathways in fluorite-structure binary oxides (i.e., HfO2 and ZrO2) – offering full scalability and CMOS compatibility – is not well-known. Additionally, since the discovery of antiferroelectricity in ZrO2, it has been well-recognized that the electrical characteristics associated with the field-induced phase transition in these materials can solve some of the most pressing challenges in modern microelectronics (energy efficiency, sub-Boltzmann logic technologies, memory and neuromorphic applications, etc.). While this sets the stage for post-scaling electronics, the physical origins of ferroelectricity and antiferroelectricity in ZrO2-based thin films has yet to be unanimously confirmed nor the phase transition experimentally visualized.
Significant gaps remain in our fundamental understanding of the structure-property relationships in polycrystalline HfO2/ZrO2-based FE/AFE thin films. Since polarization correlates with crystal structure, the application of an electric field alters the microscopic features, e.g. grain orientation, phase, size, and sub-grain characteristics (interphase boundaries and domain walls) of these materials. This complex field-induced evolution of microstructure enables electrical characteristics such as multi-level cell capabilities for embedded non-volatile memory, analog synapses, and abrupt transitions for artificial neurons. On the other hand, such evolution of microstructure poses significant challenges to performance including cycle-to-cycle and device-to-device variation, reliability, and endurance. Due to the end of dimensional scaling of transistors, materials innovation is more crucial now than ever before to the advancement of microelectronics and modern computing.
The goal of the work presented in this thesis is to advance our fundamental understanding of the structure-process-performance relationships in ferroelectric and antiferroelectric ZrO2-based thin films by establishing a foundation for high-throughput, automated microstructural analysis via a synergistic advanced microscopy characterization approach, thereby providing significant insight into processes necessary to optimize these material properties and enhance device performance while reducing power consumption in post-scaling electronics. Here, a variety of advanced microscopy characterization techniques are employed, and sample preparation procedures developed for the purpose of characterizing and quantifying the microstructural properties associated with material performance as a function of processing for polycrystalline zirconia-based thin films. These techniques include in situ TEM biasing, high resolution transmission electron microscopy (HRTEM), STEM, DF-TEM, and NBED. The first step in identifying structure-performance relationships presented here is the direct imaging of the polarization switching at the atomic and mesoscopic scales with applied bias, which suggests the presence of a field-induced phase transition with an applied field. Secondly, both local and statistical microstructural analysis of atomic structure, grain size distribution, orientation, and epitaxy are achieved and compared for both ferroelectric and antiferroelectric zirconia-based thin film capacitors, revealing structural pathways for ferroelectric phase stabilization as a function of doping and substrate. Thirdly, in addition to cross-section analysis, plan-view microstructural analysis is achieved, allowing for direct statistical quantification of real-space polycrystalline microstructure in ferroelectric zirconia-based thin films. The results from this work showcase the statistical, high-throughput characterization capabilities afforded by advanced electron microscopy for the purposes of furthering our understanding of the microstructure-property-process relationships in ferroelectric and antiferroelectric polycrystalline thin film materials.